Text orientation is classified as either positive or negative in sentiment analysis, which is viewed as a classification task. This study presents the findings of an experiment using benchmark datasets to train a sentiment classifier using Support Vector Machine (SVM). A new era that preserves the genuine nature of technology and digitalization has emerged as the globe has undergone a change. The market has changed at an astounding rate, thus it is imperative to take advantage of and inherit the benefits and prospects it offers. The emergence of web 2.0 has brought with it scalability and limitless reach, thus it would be disastrous for an organisation to ignore the new tactics in the competitive landscape this expanding virtual world has set along with its advantages. Organisations are now able to gather, classify, and analyse user evaluations and comments from microblogging sites about their services and products because to the advanced and sophisticated data mining techniques. The most classical traits were extracted using N-grams and various weighting schemes. Chi-Square weight characteristics are also investigated in order to choose informative features for the categorization. Chi-Square feature selection may significantly increase classification accuracy, according to experimental study.